Monday, April 25, 2011

Conservatives return to 151

With the long Easter weekend intervening in the election campaign, few polls have been released in the past few days. Nanos reported yesterday on a three-day roll up that excluded Friday, and they did the same today. But they won't be reporting tomorrow, as their pollsters have the day off today. Accordingly, the projection hasn't moved around much. But every polling firm should report at least once this week, so it will be a roller coaster ride.
The Conservatives are still at 38.6%, where they have been for most of the campaign. They are up one seat to 151. The Liberals are up 0.1 point to 27.5% but are down one seat to 75. The New Democrats are up 0.1 point to 19.1%, but are unchanged at 36 seats.

The Bloc Québécois is down to 8.1% and unchanged at 45 seats while the Greens are down to 5.5%.

The Conservatives gained in two key regions: half-a-point in Ontario and almost a full point in Atlantic Canada. They are now actually in the lead (by the slimmest of margins) on the East Coast.

For the Liberals, they lost in Ontario and Atlantic Canada but did make decent gains in British Columbia and Quebec. Speaking of which, Michael Ignatieff was on Tout le monde en parle last night and did well enough. He was most helped out by Dominique Michel, a well-known celebrity in Quebec (so well-known, in fact, that she and not Ignatieff was the first guest), who endorsed the Liberal leader as he was sitting right next to her. Had she said instead that she was supporting the Bloc or the NDP, it would have been devastating.

The New Democrats were stable but made another big gain in Quebec. They're now at 20.7% there and a few seats are on the tipping point, Jeanne-Le Ber in particular.

The pain continues for the Bloc, down 0.3 points to only 34.1%. They're being dragged down with every passing day and every new poll.

The one seat that changed hands was in Newfoundland & Labrador. Conservative candidate John Ottenheimer is now the projected winner in Random - Burin - St. George's over Liberal incumbent Judy Foote. Ottenheimer is a former provincial cabinet minister who retired from provincial politics in 2007.

A few other Atlantic Canadian seats are on the bubble. The gap is less than five points in the projection in Moncton - Riverview - Dieppe (Liberals lead, Conservatives trail), Saint John (CPC lead, LPC trail), West Nova (CPC lead, LPC trail), and Egmont (CPC lead, LPC trail). But if the Conservatives continue to gain while the Liberals lose, seats like Madawaska - Restigouche, Malpeque, and St. John's South - Mount Pearl could turn blue - enough to push the Tories to 154 seats in all. For the New Democrats, they might have a shot in St. John's South - Mount Pearl, South Shore - St. Margaret's, and Dartmouth - Cole Harbour. One of those seats is held by the Tories, and the first is a three-way race. So in some individual ridings out east the NDP can play a role in defeating Conservatives, along with Liberals.

Check back later today for a look at the Nanos poll and two riding polls released over the weekend in Quebec.


  1. You mention that Dominique Michel endorsed Ignatieff on Tout le Monde en parle yesterday; sorry it looked to me like a rehearsal for her fund-raising campaign for Maisonneuve-Rosemont Hospital.
    Michelle H

  2. According to the charts below, your Quebec numbers are calculated using the following weighting:

    15% - pre-election
    37% - pre-debate
    49% - post-debate

    Considering what's happening in Quebec, should most of your projection come from polls taken before the French-language debate?

    If you use only post-debate polls, the figures would be:

    31.4% - BQ
    25.9% - NDP
    19.2% - Lib
    18.3% - Cons
    3.5% - Green

    If you use only polls taken in the past week, the figures would be:

    30.5% - NDP
    28.4% - BQ
    18.5% - Lib
    17.2% - Cons
    2.9% - Green

  3. I think the piece in the Star which is linked at the end of Week 4 Ceilings tells us a great deal about what the real situation is in thre body politic.

    Disillusion, disgust, rejection, dismay.

    You name it and it's all there and that may, to an extent, explain the Advance Poll situation.

    Public is fed up with all the parties and just wants to get it over.

  4. Is it true that three of the polling companies do not include Green Party in their polls?

    Does this skew your evaluation?

  5. Nanos this morning has the Conservatives with a 13.6% lead over the Liberals and 15.6% over the NDP. If you omit the Quebec numbers the gap increases to 16% over the Liberals and 21.1% over the NDP. If the Conservatives receive 42% or more of the Rest of Canada votes they may well have a comfortable majority even if they don't get any seats in Quebec.


  6. Anonymous 10:26,

    Nanos does not prompt for any of the parties, and Ipsos does not prompt for the Greens.

    No, it shouldn't skew my evaluation as prompting for the Greens over-estimates their support, as not prompting under-estimates it. It balances out.

  7. what does 'prompt' mean in the way you use it here??

  8. Siss16,

    A poll that prompts will ask a question like "If an election were held today, would you vote for the Conservatives, the Liberals, etc." listing each of the parties.

    A poll that doesn't prompt would simply ask "Who are you going to vote for?" without naming any of the parties.

  9. "No, it shouldn't skew my evaluation as prompting for the Greens over-estimates their support, as not prompting under-estimates it. It balances out."

    Are yousure that not prompting for the greens under-estimates their support? My recollection is that Nanos didn't prompt in 2008 and it was the only pollster who was close to the actual green support (and even Nanos overstated it - although, to be fair, it's estimates of the green vote throught the campaign were pretty consistent with their actual level of support).

    If I were a betting man, I'd put money on the Green's getting closer to the levels indicated by Nanos or Ipsos than whatever is indicated by the other pollser.

  10. Why do you even give weight to pre-election polls? Things change in elections. I'm sorry, Eric, but your projection is going to be way off on election-day, unless you override your model.

  11. Hi Eric,

    Why does your daily poll average total 117% including undecided?

  12. Eric,

    An interesting trend in your federal daily averages is a significant recent increase of undecided (by 4%). On April 21 it stands at 17%. If all of them decide to show up to vote, and will go mainly with one (the most positive?) party in the day or two preceding the election day, that would be a game changer. On April 21, the support for the parties among all voters (including undecided) was 31.5% (CPC), 21.6% (Lib), 19.9% (NDP). If all 17% of undecided will go to any of these parties, the results will be quite dramatic - 48.5% for CPC (huge majority), or 38.8% for Lib (strong minority?), or 36.9% for NDP (weak NDP minority?). So it is up to all three parties to motivate those undecided to (1) vote, and (2) vote for them.

    I have a question. Have you ever analysed what happens with undecided voters, based on polling in previous elections? Does the percentage go down sharply right before the voting day? What is the percentage of those who are still undecided in the last polls, but vote? How is their vote distributed - proportional to the popular vote, or they tend to favour disproportionally a single party?

  13. Anonymous 11:44,

    Because the extra 17% is the undecided.


    I have not looked at that, but pollsters generally tell me the undecideds break in a similar proportion as the decideds.

  14. At 48% in Ontario (Nanos) the Conservatives have their majority. btw Guelph will be Tory on May 2nd. They have voted with the government for 65 years except the last election. The Conservative has split the Catholic, vote which all went to the Liberal last time. There's no way Guelph remains Liberal on May 2nd with 48% Conservative support in the province and two Catholics in the running.

  15. Well that was an interesting experience.

    Went to an Advanced Poll. This is a little town of about 6000. So what did I see??

    As the opening time approached a line formed building up to about 40 by the time the poll opened. Usual rock slow processing of course.

    All age ranges by the way. But by the time I left having waited about 25 minutes to vote that line had grown to well over 50.

    So something is going on and I suspect all the polls and polling will be for nowt !!

  16. Hi Éric,

    I'm sure you've already considered these factors (either privately or somewhere else on this blog that I haven't found), but I was just wondering what you thought of a recent 'Reality Check':

    I'm less interested in what it says about the NDP surge, and more on the methodological problems, etc. Also, do you have a sense of the expected voter turn out this election; higher or lower?

  17. I still think that a majority of voters made up their minds in December 2008.

  18. This will be an interesting one if the turnout is high as the advance polls are suggesting it will be. Obviously, as a Green, I hope many frustrated voters go to the one party that wasn't in the HOC thus the one that didn't act like a spoiled child during the past 2 years. Sadly, with the complete lack of coverage (thanks in large part to a mix of Nanos being the only daily poll and the non-invite to the debate) I wonder if we'll have any shot.

    I wonder if there is any way to start having the massive pro-Conservative coverage in the Sun papers and the massive pro-Liberal coverage in the Star count against their spending caps? Sadly, I already know the answer.

  19. Polling
    and 'the NDP surge'

    Results from four

    How can there be an almost nine-point difference in the
    Conservative vote between Ipsos Reid and Ekos?

    Or more than four points for the Greens between Ekos and
    Nanos, and more than five for the Liberals between Ipsos and Nanos?

    If the pollsters are so far apart, how can we rely on their
    interpretation of what is happening "out there" in the Canadian

    The question is important because we have reached the stage of
    the campaign where polls have become the story. The platforms have been
    released, the promises rolled out, the debates are a fading memory.

  20. I don't think anything can be extrapolated from early voter turnout. Every party could argue that that favours them. Grits can say oh those 800,000 phantom Liberals are coming out to vote this time. Conservatives can say our voters are more enthusiastic and we are more organized so the early votes are for us. Whatever. It was just a beautiful spring day and a holiday weekend on the Friday in most of the country. So people went for a walk and voted.

  21. Fascinating. Two races with which I am familiar (close by but not my riding) in which I think your analysis might be wrong: Esquimalt Juan de Fuca looks like the Conservatives' candidate will win; Troy de Souza is running like an incumbent, absent Martin, and, in Saanich Gulf Islands, it will be a closer race between the Lunn and May than you project, with her winning, if the momentum continues. Also, has their ever been an analysis of early vs. voting day voter behavior? Do early voters tend to vote the same way voting day voters do? I know in local elections it can be very different.

  22. Regarding the NDP surge:

    An effect that I would like to suggest as a possibility.

    Every 4 years in the United States they have their presidential elections and right after the nomination convention it is traditional for each of the candidates to get a "convention bounce" where (at least among adults) they temporarily enjoy a bounce of typically 12% or so in the polls....

    This bounce is typically short lived, and also much, much smaller when you narrow the sample down to likely voters who pay more attention and are thus less influenced by the assorted dog and pony shows that are US political conventions.

    How much of this Layton "bounce" is real, and how much of it is polling froth that is based in the 40 ish percent of the electorate that is quite unlikely to actually show up on election day?

    The polling being done on a daily basis for the major political parties is vastly more sophisticated that the stuff we see in the public domain, and the party leaders know the "truth" about where this election stands.

    Follow where the leaders go, and follow where they invest their ad money.

    If the leaders are in target seats that a party hopes to win, they are on offense and doing well in the polls, if they are defending party strongholds, they are on defense...


COMMENT MODERATION POLICY - Please be respectful when commenting. If choosing to remain anonymous, please sign your comment with some sort of pseudonym to avoid confusion. Please do not use any derogatory terms for fellow commenters, parties, or politicians. Inflammatory and overly partisan comments will not be posted. PLEASE KEEP DISCUSSION ON TOPIC.

Note: Only a member of this blog may post a comment.